AICAS 2026's Final Conference Program
Table of Contents
Tutorials, September 14, 2026
Tutorials: Monday, 14 September 2026 (free for all AICAS participants) | ||||
9:00:00 AM-9:15:00 AM | Opening Session | |||
9:15 AM-6:00 PM | Tutorial #1: AI-Enabled IC Design Automation and Intelligent Hardware Architectures | Tutorial #2: High-Speed Interconnect and Clock Circuits for AI Computing Infrastructures | ||
9:15 AM - 10:15 AM | Hardware-Assisted Verification for Next-Generation AI Hardware Accelerators | Introduction to High-Speed SerDes Communications | ||
10:15 AM | Tea Break (15 minutes) | |||
10:30 AM-11:30 AM | AI-Driven Shift-Left for Circuit and System Design | High-Speed Ultra-Low Latency Mixed-Signal SerDes Transceiver Design Techniques for AI Computing Infrastructures | ||
11:45 AM | Lunch | |||
1:15 PM - 2:15 PM | AI for Analog IC Design Automation: Evolution and Open Challenges | Clock Generation Techniques for High-Speed Wireline Transceivers | ||
2:15 PM - 3:15 PM | Bio-inspired Neuromorphic Olfaction: From Algorithms to Chips | Clock and Data Synchronization Techniques for High-Performance Interconnection | ||
3:15 PM | Tea Break (15 minutes) | |||
3:30 PM - 4:30 PM | Product-Ready Neuromorphic MCU at the Sensor Edge | Design of Power-Efficient Reference-less PAM4 CDR for Ultra-High-Speed Wireline Communication | ||
4:30 PM - 5:30 PM | Reconfigurable Accelerator Design for Multi-Precision Floating-Point Computing: From Processing Elements to SME-Oriented Architectures | Optical Receiver Front-Ends for AI Interconnect: Scenarios, Tradeoffs, and Design Guidelines | ||
5:30 PM - 6:00 PM | AI-Driven EDA: Reshaping Chip Design in the AI Era | | ||
Day 1, September 16, 2026
Day 1: Wednesday, 16 September 2026 | ||||
8:30 AM-5:00PM | Live demonstration & Exibits | |||
Time | Paper # | Title | Authors | Type |
Whole day | 1571274451 | Live Demonstration: A Hardware-Efficient Arc Early Detection System via DWT-ConvLSTM at Edge | Zhenhao Wang, Xuhui Du and Jialong Wang (Xidian University, China); Fan Jiang and Xian Ding (Beijing Smart-Chip Microelectronics Technology Co., Ltd); Xiaoxi Li, Xiguang Wu, Bo Li, Yan Liu, Wei Mao and Genquan Han (Xidian University, China) | Exhibit |
Whole day | 1571274667 | Live Demonstration: Edge-Deployed Compound AI with RAG for High-Accuracy Embodied Robotic Planning | Yitong Shen, Mingde Duan, Zhenhao Wang, Yingjie Zhao, Yanwen Ding and Xuhui Du (Xidian University, China); Sai Zhang (Junhe Technology (Quzhou) Co., Ltd.); Xiguang Wu, Bo Li, Yan Liu, Wei Mao and Genquan Han (Xidian University, China) | Exhibit |
Whole day | 1571306695 | Live Demonstration: A Real-Time Reservoir Processor on FPGA for Edge Temporal Prediction | Yucen Shi, Yuncheng Lu, Zehao Li, Junying Li and Tony Tae Hyoung Kim (Nanyang Technological University, Singapore) | Exhibit |
Whole day | - | Huawei exhibition | Haonan Zhang (Huawei, China) | Exhibit |
Whole day | - | CASS Vietnam Chapter | Manh-Hiep Dao, Vietnam National University, Hanoi | Exhibit |
9:00 AM - 9:30 AM | Conference Opening & GC/TPC Reporting | |||
9:30 AM - 10:30 AM | Keynote #1: From Design Automation to Autonomous Design: AI Super Agents for Chip Development | |||
10:30 AM | Tea break | |||
11:00 AM - 12:00 PM | Keynote #2: Learning from Biology - Neuromorphic Techniques for an Intelligent Vision System: Using Obstacle Avoidance as an Example | |||
12:00 PM | Lunch | |||
1:30 PM-2:45 PM | Session #1: AI Circuits and Systems for Robotics and Automation (4 papers) | |||
Time | Paper # | Title | Authors | Type |
1:30 PM | 1571268204 | A 40 µA, 1-14 V Load-Adaptive in-Situ Neural Stimulator | Dawei Li, Yao Zhou and Ziyu Chen (South-Central Minzu University, China); Dongqin Deng (Wuhan University, China); Li Zhang (Hubei University of Technology, China); Yuan Du and Li Du (Nanjing University, China); Dongsheng Liu (Huazhong University of Science and Technology, China) | Oral |
1:45 PM | 1571267971 | GAP-KV: Granularity-Aware Pruning for Chunk-Level KV Cache Compression in Long-Context LLMs | Hakyeong Lee, Sangbeom Jeong and Hyun Kim (Seoul National University of Science and Technology, Korea (South)) | Oral |
2:00 PM | 1571268277 | Lightweight Pallet Detection and Pose Estimation Technology for Pallet Truck Automated Guided Vehicles | Jize Liu, Zherong Liu (University of Electronic Science and Technology of China (UESTC), China); Tianyi Zhao (Chengdu Ruixinxing Technology Co., Ltd., China); Wentao Fan (University of Electronic Science and Technology of China (UESTC), China); Jiayao Huang (Chengdu Ruixinxing Technology Co., Ltd., China); Hang Chen, Haoxiang Long and Bozhang Hua (University of Electronic Science and Technology of China (UESTC), China); Yu Long and Ling Xu (Chengdu Ruixinxing Technology Co., Ltd., China); Ye Liu and Jun Zhou (University of Electronic Science and Technology of China (UESTC), China); | Oral |
2:15 PM | 1571274040 | Embedded AI Architecture Evaluation for in-Barn Real-Time Bovine Mastitis Detection at the Edge | Francois Gonothi Toure, Abdoulaye Baniré Diallo and Mounir Boukadoum (Université du Québec à Montréal, Canada) | Oral |
1:30 PM-2:45 PM | Session #2: AI for Circuits and Systems I (5 papers) | |||
Time | Paper # | Title | Authors | Type |
1:30 PM | 1571252614 | Beyond Iterative Search: Intelligent Generative GATv2 Framework for Analog Sizing | Fan Hu, Shaik Jani Babu and Xinfei Guo (Shanghai Jiao Tong University, China) | Oral |
1:45 PM | 1571264263 | Curriculum Learning as an Enabling Methodology for Reinforcement Learning in Analog Circuit Sizing | Douglas Mateus Machado, Nathan Chanez and Severin Trochut (STMicroelectronics, France); Paul Chollet (Télécom Paris, France); Roberto Guizzetti and Pascal Urard (STMicroelectronics, France); Van-Tam Nguyen (Télécom Paris, France) | Oral |
2:00 PM | 1571268026 | 20 Hours from Spec to Schematic: A Fully Automated Design Framework for Continuous-Time Sigma-Delta Modulators | Yuxuan Zhang, Chen Wang, Zhuohua Liu and Yuanqi Hu (Beihang University, China); Wei W. Xing (University of Sheffield, United Kingdom (Great Britain)) | Oral |
2:15 PM | 1571268245 | PIECE: Projection-Invariant Normalization for Energy-Aware Compression in Efficient Transformer Inference on Chiplets | Tzu-Ching Yang, Jiing-Ping Wang and An-Yeu (Andy) Wu (National Taiwan University, Taiwan) | Oral |
2:30 PM | 1571268458 | Multimodal Graph-Image Reinforcement Learning for VLSI Macro Placement | Yi-Syuan Shih (National Central University, Taiwan); Yu-Guang Chen (National Tsing Hua University, Taiwan); Jia-Sin Pan (National Central University, Taiwan); Shih-Cheng Huang, Cheng-Hong Tsai and De-Shiun Fu (Global Unichip Corporation, Taiwan); Mango Chia-Tso Chao (National Yang Ming Chiao Tung University, Taiwan) | Oral |
1:30 PM-2:45 PM | Session #3: Emerging applications: IoT, healthcare, smart factories, environment, robotics, autonomous vehicles, drones, AR/VR, metaverse I (4 papers) | |||
Time | Paper # | Title | Authors | Type |
1:30 PM | 1571268406 | PanLUNA: An Efficient and Robust Query-Unified Multimodal Model for Edge Biosignal Intelligence | Marija Zelic; Anna Tegon; Yawei Li; Thorir Mar Ingolfsson; Luca Benini (ETH Zürich, Switzerland) | Oral |
1:45 PM | 1571250142 | MAPLE: Efficient Agglayer Proving via off-Chain Pre-Verification for Metaverse Multi-Chain Systems | Huyen Nguyen, Lan Ngoc Le, Khoa Tan VO, Thu Nguyen and Tu-Anh Nguyen-Hoang (University of Information Technology - VNU-HCM, Vietnam) | Oral |
2:00 PM | 1571264335 | Adaptive Routing for Efficient Diffusion Transformer-Based PNI Prediction | Youngung Han (Seoul National University, Korea (South) & OUTTA, Korea (South)); Dohyun Kweon (Kyung Hee University, Korea (South)); Kyeonghun Kim (OUTTA, Korea (South)); Hyunsu Go, Jina Jeong and Suah Park (Seoul National University, Korea (South)); Induk Um (Chung-Ang University, Korea (South)); Junga Kim, Anna Jung and Yului Jeong (Seoul National University, Korea (South)); Sungha Park (Seoul National University School of Medicine, Korea (South) & Seoul National University, Korea (South)); Jinyong Jun (Seoul National University, Korea (South)); Pa Hong (Samsung Changwon Hospital, Korea (South)); Woo Kyoung Jeong (Samsung Medical Center, Korea (South)); Won Jae Lee (Samsung Medical Center, Sungkyunkwan University School of Medicine, Korea (South)); Ken Ying-Kai Liao (NVIDIA, Taiwan); Hyuk-Jae Lee and Nam-Joon Kim (Seoul National University, Korea (South)) | Oral |
2:15 PM | 1571264882 | HLCFN-PAD: Hybrid Loss-CBAM Model for Partial Fingerprint anti-Spoofing | Dao-Heng Luo and Ching-Te Chiu (National Tsing Hua University, Taiwan); Ned Lin, Ang Yu, Paul Lai and Daniel Chen (FocalTech Electronics Co., Ltd., Taiwan); Shao-En Lin (National Tsinghua University, Taiwan) | Oral |
1:30 PM-2:45 PM | Session #4: Circuits and systems for AI I (5 papers) | |||
Time | Paper # | Title | Authors | Type |
1:30 PM | 1571263159 | PHSim: An Efficient Simulator for LLM Inference on PIM-Based Heterogeneous Systems | Yueqi Zhang (Nanjing University, China); Chao Fang (KU Leuven, Belgium); Xianshi Jiang, Jie Chen and Yichuan Bai (Nanjing University, China); Peng Qin, Dawei Wang and Xu Zhang (Intel Labs, China); Yuan Du and Li Du (Nanjing University, China) | Oral |
1:45 PM | 1571265653 | AP-Rec: Adaptive Dual-Path Inference for Scalable Multi-GPU Recommendation Systems | Jiyou Hwang, Jongmin Seok and Chae Eun Rhee (Hanyang University, Korea (South)) | Oral |
2:00 PM | 1571267418 | An Accurate Yet Compact Softmax Hardware Implementation Optimized for Microscaling Data Format | Meng-Hsun Hsieh and Juinn-Dar Huang (National Yang Ming Chiao Tung University, Taiwan) | Oral |
2:15 PM | 1571268062 | A Lightweight Accelerator for Kolmogorov-Arnold Transformers in Edge Inference | Chengxing Fan, Xuliang Yu and Jie Hu (Zhejiang University, China); Wei-Han Yu (University of Macau, Macao); Liang Zhao (Zhejiang University, China) | Oral |
2:30 PM | 1571268419 | AMBER: Thermal-Aware Spatio-Temporal Scheduling for Edge LLM Inference on 3D DRAM-on-NPU Systems | Xuanda Lin, Siyao Jia and Haozhe Zhu (Fudan University, China); Xinru Jia (CiMicro.Ai, China); Shiwei Liu (Tongji University); Chixiao Chen (Fudan University, China) | Oral |
2:45 PM | Tea break | |||
3:15 PM - 4:15 PM | Session #5: Computer vision algorithms and architectures I (4 papers) | |||
Time | Paper # | Title | Authors | Type |
3:15 PM | 1571254629 | ADORE-VLM: Attention-Dynamics-Oriented Re-Entry-Based Early Exiting for Accelerating Vision-Language Model Inference | Tackwoo Nam, Gilha Lee and Hyun Kim (Seoul National University of Science and Technology, Korea (South)) | Oral |
3:30 PM | 1571264329 | Anatomy-Privileged Distillation with Token Routing for MRI-Based Prediction of Perineural Invasion | Hyunsu Go (Seoul National University, Korea (South)); Youngung Han (Seoul National University, Korea (South) & OUTTA, Korea (South)); Kyeonghun Kim (OUTTA, Korea (South)); Junga Kim (Seoul National University, Korea (South)); Dohyun Kweon (Kyung Hee University, Korea (South)); Jinyong Jun (Seoul National University, Korea (South)); Sungha Park (Seoul National University School of Medicine, Korea (South) & Seoul National University, Korea (South)); Anna Jung (Seoul National University, Korea (South)); Induk Um (Chung-Ang University, Korea (South)); Yului Jeong, Suah Park, Jina Jeong and Suemin Yang (Seoul National University, Korea (South)); Pa Hong (Samsung Changwon Hospital, Korea (South)); Woo Kyoung Jeong (Samsung Medical Center, Korea (South)); Won Jae Lee (Samsung Medical Center, Sungkyunkwan University School of Medicine, Korea (South)); Ken Ying-Kai Liao (NVIDIA, Taiwan); Hyuk-Jae Lee and Nam-Joon Kim (Seoul National University, Korea (South)) | Oral |
3:45 PM | 1571264333 | SpikeDS: Dual Sparsity Spikformer for Perineural Invasion Prediction in 3D MRI | Induk Um (Chung-Ang University, Korea (South)); Youngung Han (Seoul National University, Korea (South) & OUTTA, Korea (South)); Kyeonghun Kim (OUTTA, Korea (South)); Yului Jeong, Jina Jeong and Hyunsu Go (Seoul National University, Korea (South)); Dohyun Kweon (Kyung Hee University, Korea (South)); Sungha Park (Seoul National University School of Medicine, Korea (South) & Seoul National University, Korea (South)); Junga Kim, Anna Jung, Suah Park and Hyuk-Jae Lee (Seoul National University, Korea (South)); Pa Hong (Samsung Changwon Hospital, Korea (South)); Woo Kyoung Jeong (Samsung Medical Center, Korea (South)); Won Jae Lee (Samsung Medical Center, Sungkyunkwan University School of Medicine, Korea (South)); Ken Ying-Kai Liao (NVIDIA, Taiwan); Nam-Joon Kim (Seoul National University, Korea (South)) | Oral |
4:00 PM | 1571267460 | Robust Image Classification Under Complex Illumination: A FastViT Framework with Sharpness-Aware Minimization | Bohan Guo, Chaoyao Shen, Yuning Ji, Yu Liang and Meng Zhang (Southeast University, China) | Oral |
3:15 PM - 4:15 PM | Session #6 Special Session: AI-Powered Electronic Design Automation for IC Design I (4 papers) | |||
Time | Paper # | Title | Authors | Type |
3:15 PM | 1571266792 | SafeRTL: IP-Preserving Verilog Optimization via Principle-Guided Local-Cloud LLM Framework | Jing Wang (Nanjing University of Posts and Telecommunications, China); Zheng Li (Fudan University, China); Lei Li (The University of Hong Kong, Hong Kong); Fan He and Liyu Lin (Fudan University, China); Yao Lai (The University of Hong Kong, Hong Kong); Jiafei Yao (Nanjing University of Posts and Telecommunications, China); Yan Li and Xiaoyang Zeng (Fudan University, China); Yufeng Guo (Nanjing University of Posts and Telecommunications, China) | Oral |
3:30 PM | 1571268118 | Efficient Machine Learning-Driven RF Post-Layout Prediction via Bidirectional State-Space Vision Mamba | Hoang Trong Nguyen, Lihong Zhang and Trung Q. Duong (Memorial University of Newfoundland, Canada) | Oral |
3:45 PM | 1571268248 | AI Chip Design Tool with Self-Learning Capability for FinFET Advanced Processes | Dan Liu, Qian Wu and Xiaoshi Zhao (Beijing Tango Software Technology Company, China) | Oral |
4:00 PM | 1571268363 | Chiplet-NSRL: A Neuro-Symbolic Reinforcement Learning Approach for Interpretable Design Space Exploration of Chiplet-Based AI Accelerator | Ruifeng Zhang, Bo Li, Xiguang Wu and Wei Mao (Xidian University, China); Guoyong Shi (Shanghai Jiao Tong University, China); Yan Liu and Genquan Han (Xidian University, China) | Oral |
3:15 PM - 4:00 PM | Session #7: AI for Future Wireless Communications (2 papers) | |||
Time | Paper # | Title | Authors | Type |
3:15 PM | 1571268896 | Digital Predistortion Design with Lightweight Variation-Aware Network for Sub-THz Band | Tian Sheuan Chang and Tzu-Chia Tseng (National Yang Ming Chiao Tung University, Taiwan) | Oral |
3:30 PM | 1571272151 | A Binary Matrix Optimization Framework for the Automated Design of Broadband Microstrip Resonators | Fu Qi Chua (Universiti Malaya, Malaysia); Chao Wang and Jinghua Zhang (Shanghai Jiao Tong University, China); Kian Kit Cheah and Harikrishnan Ramiah (Universiti Malaya, Malaysia); Yongfu Li (Shanghai Jiao Tong University, China) | Oral |
3:15 PM - 4:15 PM | Session #8 Special Session: Emerging Energy-Efficient Hardware and Computing Paradigms for Edge Intelligence I (4 papers) | |||
Time | Paper # | Title | Authors | Type |
3:15 PM | 1571268395 | A Novel Delta-Sigma Computing-in-Memory Macro Based on FeRAM for Edge Computing | Huanning Zhang, Bo Li, Xiguang Wu, Wei Mao, Xiao Yu, Bing Chen, Yan Liu and Genquan Han (Xidian University, China) | Oral |
3:30 PM | 1571268472 | SpikeLogBERT: Energy-Efficient Log Parsing Using Spiking Transformer Networks | Thuan Bui and Duong Do (Swinburne Vietnam, FPT University, Vietnam); Tung Vu (Posts and Telecommunications Institute of Technology, Vietnam); Duc-Tho Mai (Academy of Cryptography Techniques, Vietnam); Cong-Kha Pham (University of Electro-Communications (UEC), Japan) | Oral |
3:45 PM | 1571268951 | A Real-Time Vietnamese Speech Acquisition and Recognition System for Control Applications on FPGA Technology | Dang-Quang Nguyen, The-Anh Nguyen, Duy-Hieu Bui and Xuan-Tu Tran (Information Technology Institute, Vietnam National University, Hanoi) | Oral |
4:00 PM | 1571273751 | MRTPE: A 28nm 131.76 TOPS/W and 8 Core RISC-V DNN Inference Processor with Multiple RISC-V Transformational Mechanism | Zhou Wang (Imperial Global Singapore, Imperial College London); Jiuren Zhou and Xiguang Wu (Xidian University, China); Xiaonan Tang (Wisemaytech, China); Shushan Qiao (Institute of Microelectronics of the Chinese Academy of Sciences, China); Tianchun Ye (IME-CAS, China); Anil A Bharath (Imperial College & Cortexica Vision Systems, United Kingdom (Great Britain)); Emm Mic Drakakis (Imperial College London, United Kingdom (Great Britain)) | Oral |
4:15 PM | Tea break | |||
4:30 PM-5:45 PM | Session #9: Architecture for AI computing I (5 papers) | |||
Time | Paper # | Title | Authors | Type |
4:30 PM | 1571264636 | KV Cache Management for Block Diffusion Language Models on Multi-Level NV-Memory | Masafumi Higashi, Shota Suzuki, Yuya Yamamoto, Naoko Misawa, Chihiro Matsui and Ken Takeuchi (The University of Tokyo, Japan) | Oral |
4:45 PM | 1571268263 | A Low-Power Sparse Convolution Accelerator with Idle-First-Task-Assignment for Edge Vision | Jingyue Zhuge (TUD Dresden University of Technology, Germany & ScaDS.AI, Germany); Johannes Partzsch (TUD Dresden University of Technology, Germany); Christian Mayr (TU Dresden, Germany) | Oral |
5:00 PM | 1571268280 | Sub-100KB on-Device CNN Training via Dynamic Sparse Updates and Hardware--Software Co-Design | Tian Sheuan Chang and I-Hsuan Li (National Yang Ming Chiao Tung University, Taiwan) | Oral |
5:15 PM | 1571268292 | A Multi-Precision Accelerator for Partial Differential Equations Solving with Collaborative Physics-Informed Neural Networks and Finite Difference Methods | Yichuan Bai, Botao Wu and Xiaoyu Chang (Nanjing University, China); Chongkang Tan (Ant group, China); Yuan Du and Li Du (Nanjing University, China) | Oral |
5:30 PM | 1571274252 | When NPUs are Not Always Faster: A Stage-Level Analysis of Mobile LLM Inference | Pu Li, Jiawen Qi and Qinyu Chen (Leiden University, The Netherlands) | Oral |
4:30 PM-5:45 PM | Session #10: Hardware/Software Co-Design and Design Automation for AI systems I (5 papers) | |||
Time | Paper # | Title | Authors | Type |
4:30 PM | 1571265285 | DISA: Algorithm-Hardware Co-Design of a Sparsity-Induced Deep Unfolding Network for Edge Compressive Sensing | Xuewei Cao, Wenbiao Zhou, Fengchi Lu, Wenxuan Niu and Zhihan Zhang (Beijing Institute of Technology, China) | Oral |
4:45 PM | 1571266794 | SHIELD: A Segmented Hierarchical Memory Architecture for Energy-Efficient LLM Inference on Edge NPUs | Jintao Zhang and Xuanyao Fong (National University of Singapore, Singapore) | Oral |
5:00 PM | 1571267994 | NPN-ApproxLib: DNN-Aware Approximate Multiplier Design via NPN Classes | Sallar Ahmadi-Pour, Chandan Kumar Jha and Sajjad Parvin (University of Bremen, Germany); Rolf Drechsler (University of Bremen/DFKI, Germany) | Oral |
5:15 PM | 1571268453 | DiffABCO: Shift-Left Wafer-Scale Integration Design via Differentiable Architecture-Backend Co-Optimization | Zhen Zhuang (Chinese University of Hong Kong, Hong Kong); Zheng Yang (Georgia Institute of Technology, USA); Kai-Yuan Chao (Siemens Digital Industries Software, USA); Tsung-Yi Ho (The Chinese University of Hong Kong, Hong Kong) | Oral |
5:30 PM | 1571268123 | SNN2Bitstream: Automated Model-to-FPGA HLS-Based Framework for Spiking Neural Network Hardware Accelerators | Xindan Zhang, Ioannis Kaskampas, Andrea Pinna and Haralampos-G. Stratigopoulos (Sorbonne Université, CNRS, LIP6, France) | Oral |
4:30 PM-5:45 PM | Session #11: Emerging computing devices and materials for AI (6 papers) | |||
Time | Paper # | Title | Authors | Type |
4:30 PM | 1571268285 | Dual-Channel-Shared OFA-p-SOT-MRAM CIM Macro for High-Throughput Binary Similarity Search in LLM KV-Cache Compression | Fantao Gao, Cancheng Xiao, Dingsong Jiang, Xun Tang and Jiahao Zhao (Tsinghua University, China); Tao Yang (ENFI Engineering Corporation, China); Jianshi Tang (Tsinghua University, China); Tianxiang Nan (School of Integrated Circuits, Tsinghua University, China) | Oral |
4:45 PM | 1571274339 | Neural Sequence Models for Memristor State Variable Modelling and Prediction | Rishabh Mallik and Ming-Jay Yang, Stephan Menzel, Emre Neftci, John Paul Strachan (Peter Grunberg Institute, Forschungszentrum Julich, Germany) | Oral |
5:00 PM | 1571264447 | TinyGLASS: Real-Time Self-Supervised in-Sensor Anomaly Detection | Pietro Bonazzi (ETH Zurich, Switzerland); Rafael Sutter (ETH Zürich, Switzerland); Luigi Capogrosso (Interdisciplinary Transformation University of Austria, Austria); Mischa Buob (Swiss Engineering Partners AG, Switzerland); Michele Magno (ETH Zurich, Switzerland) | Oral |
5:15 PM | 1571268343 | A Transposable VC-SOT-MRAM CIM Macro for Edge AI Tasks with High Linearity | Bingqian Song, Cancheng Xiao, Yiyuan Xiao, En Li, Jiahao Zhao, Yining Li, Yuheng Yang and Jianshi Tang (Tsinghua University, China); Qi Hu (Beijing Superstring Academy of Memory Technology, China); Tianxiang Nan (School of Integrated Circuits, Tsinghua University, China) | Oral |
5:30 PM | 1571268494 | A Hardware-Oriented Event-Driven Gesture Recognition System with a Geometry-Token Early-Exit Classifier for a Perovskite Vision Platform | Junying Li and Chufeng Yang (Nanyang Technological University, Singapore); Kyung-Chan An (Seokyeong University, Korea (South)); Yuncheng Lu and Kiho Seong (Nanyang Technological University, Singapore); Ng Si En Timothy (Stanford University, USA); Shibi Varku, Natalia Yantara and Qinjie Wu (Nanyang Technological University, Singapore); Arindam Basu (City University of Hong Kong, Germany); Nripan Mathews and Tony Tae Hyoung Kim (Nanyang Technological University, Singapore) | Oral |
5:45 PM | 1571268512 | 3D Monolithic Integrated SNN for Ultra-Low Energy ECG Arrhythmia Classification | Shreesh Shrinivas Nagral (Indian Institute of Technology, Bombay, India); Harshitha Gangu, Laxmeesha Somappa and Veeresh Deshpande (Indian Institute of Technology Bombay, India) | Oral |
4:30 PM-5:45 PM | Session #12: In-Memory Computing I (5 papers) | |||
Time | Paper # | Title | Authors | Type |
4:30 PM | 1571265000 | Analysis of Accuracy, Precision, and Error Sources in Bitline Voltage Regulator for AIMC Accelerators | Berk Bozkurt and Riccardo Vignali (University of Pavia, Italy); Riccardo Zurla (STMicroelectronics, Pavia, Italy); Marco Pasotti (STMicroelectronics, Agrate Brianza, Italy); Andrea Lico and Alessio Antolini (ARCES-DEI, University of Bologna, Italy); Eleonora Franchi (University of Bologna, Italy); Alessandro Cabrini (University of Pavia, Italy) | Oral |
4:45 PM | 1571266241 | A 15.7TOPS/W NAND Flash-Based Process-in-Memory Featuring a Time-Domain Computation Unit | Hyuk-Jin Lee and Kyunghoon Pyo (Korea Advanced Institute of Science & Technology, Korea (South)); SeongHwan Cho (Korea Advanced Institute of Science and Technology (KAIST), Korea (South)) | Oral |
5:00 PM | 1571267829 | NV-Memory Centric Liquid Neural Network Accelerator for Reflexive Robotic Control | Yuya Yamamoto, Naoko Misawa, Chihiro Matsui and Ken Takeuchi (The University of Tokyo, Japan) | Oral |
5:15 PM | 1571268180 | PSD: Pairwise Signed-Digit for SRAM-Based Digital Compute-in-Memory with OR-Approximated Adder Tree | Jiing-Ping Wang and An-Yeu Wu (National Taiwan University, Taiwan) | Oral |
5:30 PM | 1571268234 | An Energy-Efficient Multi-Edge Multi-Input Delay Cell with Asymmetric Parasitic Compensation for TD-CIM | Junyeong Heo, Joo Hyeon Jeon, Sung In Cho and Jinwook Burm (Sogang University, Korea (South)) | Oral |
7:00 PM | Welcome reception | |||
Day 2, September 17, 2026
Day 2: Thursday, 17 September 2026 | ||||
8:30 AM-5:00PM | Live demonstration & Exibits | |||
Time | Paper # | Title | Authors | Type |
Whole day | 1571274451 | Live Demonstration: A Hardware-Efficient Arc Early Detection System via DWT-ConvLSTM at Edge | Zhenhao Wang, Xuhui Du and Jialong Wang (Xidian University, China); Fan Jiang and Xian Ding (Beijing Smart-Chip Microelectronics Technology Co., Ltd); Xiaoxi Li, Xiguang Wu, Bo Li, Yan Liu, Wei Mao and Genquan Han (Xidian University, China) | Exhibit |
Whole day | 1571274667 | Live Demonstration: Edge-Deployed Compound AI with RAG for High-Accuracy Embodied Robotic Planning | Yitong Shen, Mingde Duan, Zhenhao Wang, Yingjie Zhao, Yanwen Ding and Xuhui Du (Xidian University, China); Sai Zhang (Junhe Technology (Quzhou) Co., Ltd.); Xiguang Wu, Bo Li, Yan Liu, Wei Mao and Genquan Han (Xidian University, China) | Exhibit |
Whole day | 1571306695 | Live Demonstration: A Real-Time Reservoir Processor on FPGA for Edge Temporal Prediction | Yucen Shi, Yuncheng Lu, Zehao Li, Junying Li and Tony Tae Hyoung Kim (Nanyang Technological University, Singapore) | Exhibit |
Whole day | - | Huawei exhibition | Haonan Zhang (Huawei, China) | Exhibit |
Whole day | - | CASS Vietnam Chapter | Manh-Hiep Dao, Vietnam National University, Hanoi | Exhibit |
9:00 AM - 10:00 AM | Keynote #3: From FLOPS to Intelligence per Joule: The Case for Adaptive Dataflow | |||
10:00 AM | Tea break | |||
10:30 AM - 11:30 AM | Keynote #4: From Intelligence to Infrastructure: Building the Global Ecosystem for the Next AI Era | |||
11:45 AM | Lunch | |||
1:15 PM - 2:30 PM | Session #13: Architecture for AI computing II (5 papers) | |||
Time | Paper # | Title | Authors | Type |
1:15 PM | 1571268455 | Area- and Energy-Efficient Base-2 Softmax Design for DNN Inference with 1.95x Scaled Efficiency | Ting-Yi Chan, Hsiao-Jung Lin, Min-Yu Lin, Nidhi and Jinn-Shyan Wang (National Chung Cheng University, Taiwan) | Oral |
1:30 PM | 1571268928 | CRISP: Partitioning LLM Inference Across Main Memory and HBM via Input Sparsification and Low-Rank Approximation | Koki Shibata and Shota Suzuki (The University of Tokyo, Japan); Naoko Misawa (The University of Tokyo); Chihiro Matsui (The University of Tokyo, Japan); Ken Takeuchi (The University of Tokyo) | Oral |
1:45 PM | 1571273686 | SARA: A Scalable ASIP-Based RISC-V Architecture for Edge AI Applications | Kuan-Wei Chen, Yu-Hao Cheng, Yu-Hsiang Lee and Hsie-Chia Chang (National Chiao Tung University, Taiwan) | Oral |
2:15 PM | 1571274664 | Low-Power GELU-Induced Dynamic Precision for Feed-Forward Networks in Transformers | Kai-Chieh Hsu and Tian Sheuan Chang (National Yang Ming Chiao Tung University, Taiwan) | Oral |
1:15 PM - 2:30 PM | Session #14: AI for Circuits and Systems II (5 papers) | |||
Time | Paper # | Title | Authors | Type |
1:15 PM | 1571268460 | Distributed VLM and CNN Co-Processing for Edge AI over Non-Equalized RFSoC DAC-to-ADC Raw Channels | Muhammad Junaid, Dohyeon Kwon and Eva Cha (Daegu Gyeongbuk Institute of Science and Technology, Korea (South)); Korkut K Tokgöz (Sabanci University, Turkey & Evrim Co. Ltd., Japan); Gain Kim (DGIST, Korea (South)) | Oral |
1:30 PM | 1571268502 | Toward Resource-Efficient Spiking Neural Networks for MRI Classification | Khoa-Sang Nguyen and Quang-Kien Trinh (Le Quy Don Technical University, Vietnam); Deepu John (University College Dublin, Ireland); Dinh-Toan Le (108 Military Central Hospital, Vietnam); Thi-Thuy-Linh Le and Thi-Hong-Tham Tran (Le Quy Don Technical University, Vietnam) | Oral |
1:45 PM | 1571274172 | Machine Learning-Assisted Design and Optimization of Tunable Bandpass LNA for Ultrasound Receiver Analog-Frontend | Xie Xinzhe (Nanyang Technological University, Singapore); Jinhai Hu (Institute of Microelectronics, Agency for Science, Technology and Research, Singapore); Zhaoyang Cao, Siqi Zhang, Jiaju Lu and Wang Ling Goh (Nanyang Technological University, Singapore); Liter Siek (Nanyang Technological University & Deputy Director, VIRTUS-IC Design Centre of Excellence, Singapore); Abdulkadir C Yucel (Nanyang Technological University, Singapore); Yuan Gao (Institute of Microelectronics, A*STAR, Singapore) | Oral |
2:00 PM | 1571274661 | InvCal-ITF: Fast and Accurate ITF Calibration via Surrogate-Assisted Inverse Optimization | Tianyi Xu and Hongwei Yao, Bingyi Ye and Yang Shen, Xiaojin Li, Yabin Sun and Yuhang Zhang (East China Normal University, China) | Oral |
2:15 PM | 1571274669 | MonarchGRU: A Low-Footprint Recurrent Neural Network for Audio Keyword Spotting System | Soujanya Bhowmick (Aalto University, Finland); Sangyeob Kim (Yonsei University, Korea (South)); Qinyu Chen (Leiden University, The Netherlands); Chang Gao (Delft University of Technology, The Netherlands); Kwantae Kim (Aalto University, Finland) | Oral |
1:15 PM - 2:30 PM | Session #15 Circuits and systems for AI II (5 papers) | |||
Time | Paper # | Title | Authors | Type |
1:15 PM | 1571268251 | CMOS Oscillator Cluster with Analog-Tunable Coupling for Frequency-Encoded Computing | Clément Dufossée, Jean-Baptiste Begueret, Thierry Taris, Sylvain Saïghi, François Sandrez and Adrien F. Vincent (University of Bordeaux, France) | Oral |
1:30 PM | 1571268488 | HetKey: A Cross-Device Benchmark and Edge AI Pipeline for Acoustic Keystroke Attacks | Tuan-Duc Nguyen and Thuan Bui (Swinburne Vietnam, FPT University, Vietnam); Dinh-Dung Nguyen (Swinburne Vietnam, FPT University, Hanoi, Vietnam); Tung Vu (Posts and Telecommunications Institute of Technology, Vietnam); Long Pham and Duc-Tho Mai (Academy of Cryptography Techniques, Vietnam); Cong-Kha Pham (University of Electro-Communications (UEC), Japan) | Oral |
1:45 PM | 1571273620 | Hardware Acceleration of Layer Normalization on SoC for Artificial Intelligence on the Edge | Ettore Napoli and Efstratios Zacharelos (University of Salerno, Italy) | Oral |
2:00 PM | 1571273719 | A RISC-V Based Hardware Platform with CNN Acceleration for Real-Time Neonatal Seizure Detection | Lakshmi Iyer (Indian Institute of Technology, Bombay, India); Ankur Gupta, Shrey Aggarwal, Mithin K and Laxmeesha Somappa (Indian Institute of Technology Bombay, India) | Oral |
2:15 PM | 1571274113 | High-Reliability O(1) Logic-Depth LDPC Decoder for Edge AI Storage | Yanling Jiang, Junyi Wang, Xihui Qian, Zhixiong Li, Yida Li and Longyang Lin (Southern University of Science and Technology, China) | Oral |
1:15 PM - 2:30 PM | Session #16 Special Session: Specialized Hardware and Systems for Edge Large Language Model Acceleration and Application, Precision-Scalable and Reliability-Aware AI Circuits for Closed-Loop Edge Autonomy, Multimedia Systems and Applications for AI (4 papers) | |||
Time | Paper # | Title | Authors | Type |
1:15 PM | 1571273554 | A Low-Latency and Low-Hallucination Edge Compound AI System via RAG for Embodied Robotics | Mingde Duan, Yitong Shen, Zhenhao Wang, Yingjie Zhao, Yanwen Ding and Xuhui Du (Xidian University, China); Sai Zhang (Junhe Technology (Quzhou) Co., Ltd.); Xiguang Wu, Bo Li, Yan Liu, Wei Mao and Genquan Han (Xidian University, China) | Oral |
1:30 PM | 1571273607 | OPAL: Operator-Aware Precision-Sparsity Co-Scheduling for CIM-Based Transformer Inference | Yitong Shen, Can Liu and Yukun Fan (Xidian University, China); Jie Gan and Zhe Wang (China); Xin Jin (Beijing Smartchip Microelectronics Technology Company Limited, China); Xiguang Wu, Bo Li, Yan Liu, Wei Mao and Genquan Han (Xidian University, China) | Oral |
1:45 PM | 1571274655 | Stochastic-LLM: Exploring Sign-Magnitude Stochastic Computing for Low-Power Edge LLM Inference on FPGAs | Yanwen Ding (Xidian University, China); Shirui Zhao (HKUST, Hong Kong); Cheng Zhang, Mingde Duan, Yitong Shen, Xuhui Du, Xiguang Wu, Bo Li, Yan Liu, Wei Mao and Genquan Han (Xidian University, China) | Oral |
2:00 PM | 1571276533 | CIMBA: Hybrid State-Space Model Inference with Precision-Adaptive Hybrid Compute-in-Memory | Avadh Rajesh Harkishanka, Jintao Zhang, Yimin Wang and Xuanyao Fong (National University of Singapore, Singapore) | Oral |
2:30 PM | Tea break | |||
3:00 PM - 4:00 PM | Session #17: Computer vision algorithms and architectures II (3 papers) | |||
Time | Paper # | Title | Authors | Type |
3:00 PM | 1571267969 | On the Accuracy Flip of Vision Transformers Under Ultra-Low-Bit Post-Training Quantization | Hyunha Hwang, Hyuk-Jae Lee and Xuan Truong Nguyen (Seoul National University, Korea (South)) | Oral |
3:15 PM | 1571268473 | LSTA-Net: A Spatiotemporally Decoupled Network with Single-Query Transformer for Low Power Event-Based Eye Tracking | Shangyu Yang, Zhijie Jian and Yanbo Lin (Tsinghua University, China); Hong Chen (Tsinghua, China) | Oral |
3:30 PM | 1571273530 | ASIC Functional Cell-Level Classification on Infra-Red, in-Situ (IRIS) Data for Hardware Assurance | Kakuly Mittal (Nanyang Technological University, Singapore); Andrew Huang (Baochip, Inc., Singapore); Bah Hwee Gwee (NTU, Singapore) | Oral |
3:00 PM - 4:00 PM | Session #18 Special Session: AI-Powered Electronic Design Automation for IC Design II (4 papers) | |||
Time | Paper # | Title | Authors | Type |
3:15 PM | 1571270399 | AST-GWO: Adaptive Sigmoid Transfer Grey Wolf Optimizer for Improving Search Dynamics in Logic Synthesis | Dat Tan Tran, Thuc Duy Le, Gia Huy Duong and Hoang Trang (Ho Chi Minh City University of Technology, Vietnam) | Oral |
3:30 PM | 1571274086 | Component Centric Placement Using Deep Reinforcement Learning | Kart Leong Lim (Institute of Microelectronics, Singapore) | Oral |
3:45 PM | 1571274215 | Multi-Objective Reinforcement-Learning-Based Fogging and Proximity-Aware Placement Method for Analog and Mixed-Signal Circuits | Mirvala Sadrafshari, Octavia A. Dobre and Lihong Zhang (Memorial University of Newfoundland, Canada) | Oral |
4:00 PM | 1571274985 | GRPO-UVM: A Reinforcement Learning-Driven Verification Framework for AI Compute Systems | Jiaxiang Wang and Tingjie Yang (Xidian University, China); Shirui Zhao (HKUST, Hong Kong); Yingjie Zhao, Cheng Zhang, Zhenhao Wang, Jialong Wang, Xiguang Wu, Bo Li, Yan Liu, Wei Mao and Genquan Han (Xidian University, China) | Oral |
3:00 PM - 4:00 PM | Session #19 Special Session: Emerging Energy-Efficient Hardware and Computing Paradigms for Edge Intelligence II (4 papers) | |||
Time | Paper # | Title | Authors | Type |
3:15 PM | 1571273984 | Energy-Efficient SNN-Based Person Detection with Adaptive Burst Suppression | Alimatou Sadia Anike Memudu (Université Côte d'Azur, France & IMRA Europe, France); Jean Martinet (Univesité Côte d'Azur, France); Yuta Nakano (IMRA europe, France) | Oral |
3:30 PM | 1571274209 | A Novel Ferroelectric Computing-in-Memory Framework for Solving Edge Robotic Equations | Xueyang Bai, Bo Li, Xiguang Wu and Wei Mao (Xidian University, China); Guoyong Shi (Shanghai Jiao Tong University, China); Yan Liu and Genquan Han (Xidian University, China) | Oral |
3:45 PM | 1571274465 | A Hardware-Efficient DWT-ConvLSTM Spatio-Temporal System for Arc Fault Early Detection at the Edge | Xuhui Du, Yingjie Zhao, Zhenhao Wang and Jialong Wang (Xidian University, China); Mingde Duan (Hangzhou Institute of Technology, Xidian University, China); Jie Gan and Zhe Wang (China); Xin Jin (Beijing Smartchip Microelectronics Technology Company Limited, China); Xiaoxi Li, Xiguang Wu, Bo Li, Yan Liu, Wei Mao and Genquan Han (Xidian University, China) | Oral |
4:00 PM | 1571275037 | Hardware Implementation of Efficient Reconfigurable 2D FFT Processor for SAR Imaging | Wenbo Ding and Weiwei Shi, Bo Zhao (Shenzhen University, China) | Oral |
3:00 PM - 4:00 PM | Grand Challenge SS'25 | |||
Time | Paper # | Title | Authors | Type |
3:15 PM | 1571268392 | An Ultra-Efficient Inference Pipeline for Qwen2.5 on Yitian 710: Synergizing Finetuning, Quantization and Inference Optimization | Wenyuan Zhao, Ruchao Fan, Zhao Cang, Yongming Tang and Hao Sun (Southeast University, China) | Oral |
3:30 PM | 1571270318 | AICAS Grand Challenge 2025: Optimization and Application of Edge LLMs | Zhenhao Wang and Cheng Zhang, Tingjie Yang and Huanning Zhang (Xidian University, China); Guosheng Yu (T-HEAD Semiconductor Co., Ltd) Evens Pan (Arm Technology (China) Co., Ltd); Liang Chang (University of Electronic Science and Technology of China, China); Yuan Du and Li Du (Nanjing University, China); Zhezhi He and Yongfu Li (Shanghai Jiao Tong University, China); Bo Li and Wei Mao (Xidian University, China) | Oral |
3:45 PM | 1571271740 | GRASP: A Gm/Id-Based Multi-Agent Reasoning Framework for Sizing Parameters in Amplifiers | Ji Zhuang, Zishuo Wang, Jinyi Shen, Zihao Chen and Fan Yang (Fudan University, China) | Oral |
4:00 PM | 1571273651 | ELLME: A 280-Tokens/s Real-Time FPGA LLM Accelerator for End-to-End Qwen-2.5 Inference | Yujie Ma, Linqi Li, Yuman Zeng, Zhirui Huang and Chixiao Chen, Qi Liu and Haozhe Zhu (Fudan University, China) | Oral |
4:00 PM | Tea break | |||
4:15 PM - 5:45 PM | Session #20: Emerging applications: IoT, healthcare, smart factories, environment, robotics, autonomous vehicles, drones, AR/VR, metaverse II (4 papers) | |||
Time | Paper # | Title | Authors | Type |
4:15 PM | 1571268423 | Hardware-Aware Semantic Segmentation Engine for on-Chip Retinal Prosthesis Navigation | Arnav Agarwal (Indian Institute of Technology Bombay, India); Daksh Sawke (Indian Institute of Technology, Bombay, India); Laxmeesha Somappa (Indian Institute of Technology Bombay, India) | Oral |
4:30 PM | 1571268450 | Enhancing Text Entry Efficiency in AR/VR with a Pinch- and Thumb-Click Gesture-Based Virtual Keyboard | Youngmock Cho, Donggyu Kim, Viduranga Munasinghe, Tae-Ho Lee and Hyuk-Jae Lee (Seoul National University, Korea (South)) | Oral |
4:45 PM | 1571272160 | QR-Coded Microstrip Antennas: End-To-End Co-Design of URL Extensions and RF Performance for IoT and IP Protection | Yuting Hou (Shanghai Jiao Tong University, China & Xi'an Railway Vocational Technical Institute, China); Chao Wang and Jinghua Zhang (Shanghai Jiao Tong University, China); Fu Qi Chua (Universiti Malaya, Malaysia); Liyuan Du (Xi'an Railway Vocational Technical Institute, China); Mayank Kumar Singh (Shanghai Jiao Tong University, China); Guoliang Feng (Xi'an Railway Vocational Technical Institute, China); Yongfu Li (Shanghai Jiao Tong University, China) | Oral |
5:00 PM | 1571273958 | ULASC: An Ultra-Lightweight Acoustic Scene Classification Model for IoT Devices | Xiben Jiao, Yuanchang Sun, Xiaotian Wei, Fan He, Jiacheng Liang, Rui Hong, Haipeng Liu and Xingdan Cai (University of Electronic Science and Technology of China, China); Weixian Kang (Chip-Artisan Technology (Chengdu) Co. Ltd, China); Xinyi Jing, Jiahao Liu and Jun Zhou (University of Electronic Science and Technology of China, China) | Oral |
4:15 PM - 5:45 PM | Session #21: Hardware/Software Co-Design and Design Automation for AI systems II (5 papers) | |||
Time | Paper # | Title | Authors | Type |
4:15 PM | 1571268164 | Front-End Quantization Can Recover Accuracy Under Bandwidth-Induced Covariate Shift in Keyword Spotting | Nazareno Sacchi (ETHZ - ETH Zurich, Switzerland & CSEM, Switzerland); Taekwang Jang (ETHZ, Switzerland); Komail Badami (CSEM, Switzerland) | Oral |
4:30 PM | 1571268199 | Energy Efficient Analog Neural Network via SHAPE: Sensitivity-Aware Heterogeneous Power Allocation Framework | Leyang Jiang (Shanghai Jiaotong University, China); Siyun Ding, Ruge Xu, Xinfei Guo and Xuyang Lu (Shanghai Jiao Tong University, China) | Oral |
4:45 PM | 1571268380 | Performance-Aware Rank Adjustment for Hardware-Friendly Fine-Tuning in Edge AI Accelerators Exploiting Sparsity and Scalability | Bo-Yang Lin and Cheng-Han Ho (National Central University, Taiwan); Pei-Yun Tsai (National Taiwan University, Taiwan) | Oral |
5:00 PM | 1571268438 | LPAC-MoE: Lossless and Prefetch-Aware FP8 MoE Compression via Algorithm-Hardware Co-Design | Qiye Ding, Dingyang Zou and Zhongfeng Wang (Nanjing University, China) | Oral |
5:15 PM | 1571268065 | TRACE: Temporal Reuse and Communication-Aware Mapping for Heterogeneous CIM Chiplet Systems | Zixuan Wang, Abhishek Tyagi and Xuanyao Fong (National University of Singapore, Singapore) | Oral |
4:15 PM - 5:45 PM | Session #22 Special Session: Compute-in-Memory Co-Design: From Emerging Devices to AI Application Optimization (5 papers) | |||
Time | Paper # | Title | Authors | Type |
4:15 PM | 1571273940 | OpenMENA: An Open-Source Memristor Interfacing and Compute Board for Neuromorphic Edge-AI Applications | Ali Safa (Hamad Bin Khalifa University, Qatar); Farida Ali Mohsen (College of Science and Engineering, Qatar); Zainab Ali, Bo Wang and Amin Bermak (Hamad Bin Khalifa University, Qatar) | Oral |
4:30 PM | 1571274720 | A CIM-Optimized Navigation Algorithm Based on Grid Cells for Edge Applications | Mu Rongjun (Zhejiang University, China); Jiabao Ye (China); Bing Chen (Xidian University, China); Yong Ding (Zhejiang University, China) | Oral |
4:45 PM | 1571274728 | Phase-Domain Neural Computation Based on Multi-Injection Locked Oscillators | Badets Franck (CEA, France); Kamal Danouchi (CEA-Leti, France); Jean-Baptiste Casanova (CEA-List, France) | Oral |
5:15 PM | 1571274868 | Beyond Evolutionary Search: A General Reinforcement Learning Framework for Scalable Hardware Trojan Mitigation on Evolvable Hardware Platforms | Gaurav Ravindra Nayak and Niranjan Gopal and Madhav Rao (International Institute of Information Technology, Bangalore, India) | Oral |
5:30 PM | 1571274944 | Modeling of Thin-Film Transistors and Memristors from Fabricated Devices with Applications in Memristor-Based Circuit Design for Neuromorphic AI Computing | Ngoc-Lam Pham Le (Vietnam National University – Hanoi); Phu-Quan Pham (Vietnam National University-HCM); Viet-An Dang, Hoang-Viet Vu, Thuat Nguyen-Tran and The-Vinh Tran (Vietnam National University – Hanoi) | Oral |
4:15 PM - 5:45 PM | Grand Challenge SS'26 | |||
4:15 PM | 1571332291 | VersaVLM: A Phase-Specialized FPGA NPU with Log8 Attention for VLM Inference | Zhenyang Gu, Jiakun Shen, Xichen Wang, Lanqi Ma and Gengsheng Chen (Fudan University, China) | Oral |
4:30 PM | 1571334356 | Overcoming the Kernel Launch Bottleneck: 11x Faster VLM Inference with Operator Fusion and CUDA Graphs | Thomas Yeung (TWGHs Kap Yan Directors' College, Hong Kong); YiLin Zhu (East China Normal University, China) | Oral |
4:45 PM | 1571335416 | AICAS Grand Challenge 2026: Efficient VLM Inference on AI Chips and FPGA-Based Edge Systems | Cai Binglun (Shanghai Jiao Tong University, China); Linqi Li (Fudan University, China); Kang You (Shanghai Jiao Tong University, China); Bo Li and Wei Mao (Xidian University, China); Liang Chang (University of Electronic Science and Technology of China, China); Yuhang Zhang (East China Normal University, China); Yiming Gan (Institute of Computing Technology, Chinese Academy of Sciences, China); Shimeng Qian (Pingtouge Semiconductor Co Ltd, China); Yuan Du and Li Du (Nanjing University, China); Yongfu Li (Shanghai Jiao Tong University, China); Haozhe Zhu (Fudan University, China); Zhezhi He (Shanghai Jiao Tong University, China) | Oral |
5:00 PM | 1571335654 | UniVLM: Unified ViT-LLM Accelerator for Efficient End-to-End VLM Inference on FPGA | Le Zhong, Linfeng Zhong, Chunbo Li, Huifeng Wen, Yunpeng Wang and Haoyang Ping (Peking University, China); Meng Li (Peking Universitu, China); Yuan Wang (Peking University, China) | Oral |
5:15 PM | 1571335704 | End-to-End System Optimization for Low-Latency Single-Request Vision Language Model Inference | Damhyun Lee, Jongwon Lee, Yujun Heo, Sihwa Lee and Jungwook Choi (Hanyang University, Korea (South)) | Oral |
5:30 PM | 1571335705 | Memory-Centric VLM Deployment for Bandwidth-Constrained Heterogeneous Systems | Yuqi Guo (University of Electronic Science and Technology of China, China); Ruiqi Tang (Fudan University, China); Chuqi Zhao (Vanderbilt University, USA); Yucheng Sun (University of Electronic Science and Technology of China, China); Bo Yan (University of Electronic Science and Technology of China) | Oral |
5:45 PM | 1571335713 | CASA: Context-Anchored Speculative Attention and System Co-Design for High-Throughput Vision-Language Inference | Yongkang Huang, Tianjiao Yang, Shuyan Chen and Jiang Zhu (China) | Oral |
7:00 PM | Banquet & Award Ceremony | |||
Day 3, September 18, 2026
Day 3: Friday, 18 September 2026 | ||||
9:00 AM - 10:30 AM | Session #23: Neuromorphic and bioinspired circuits, processors, systems, and their applications I (5 papers) | |||
Time | Paper # | Title | Authors | Type |
9:00 AM | 1571262851 | BCPNNTorch: A PyTorch Simulation Framework for BCPNN-Based Associative Memory | Xiaoze Yan, Deyu Wang, Yu Zhuoquan and Huidong Ji (Fudan University, China); Yu Yang (KTH Royal Institute of Technology,, Sweden); Qi Jiang (Fudan University, China); Ahmed Hemani (Royal Institute of Technology, Sweden); Anders Lansner (Royal Institute of Technology (KTH), Sweden); Jiawei Xu (University of Macau, Macao); Lirong Zheng (Fudan University, China); Zhuo Zou (Fudan University & KTH Royal Institute of Technology, Sweden) | Oral |
9:15 AM | 1571267548 | Adversarial Vulnerability of Neuromorphic Processors Demonstrated on Loihi Chip | Spyridon Raptis and Haralampos-G. Stratigopoulos (Sorbonne Université, CNRS, LIP6, France) | Oral |
9:30 AM | 1571267931 | A Pulse Width Modulation-Based Compensation Circuit for Improving Linearity of Memristor Conductance Modulation | Kyusik Mun, Hyeongcheol Jo, Jaehyun Kang and Donghyuk Shin (Korea Institute of Science and Technology, Korea (South)); Joon Young Kwak (Ewha Womans University, Korea (South)); Jongkil Park, Suyoun Lee, Inho Kim, Jong-Keuk Park, Changhyuk Lee and Seongsik Park (Korea Institute of Science and Technology, Korea (South)); Hyung-Min Lee (Korea University, Korea (South)); Yeonjoo Jeong and Jaewook Kim (Korea Institute of Science and Technology, Korea (South)) | Oral |
9:45 AM | 1571267980 | A Continuous-Time QUBO Solver Using a Chain-RRAM-Based Hopfield Neural Network | Congcong Hong and Zhong Sun (Peking University, China) | Oral |
10:00 AM | 1571268029 | Hyperdimensional Computing Training and Inference with Eight-Bit Floating-Point Formats | Andrea Marcelli and Marco Angioli (Sapienza University of Rome, Italy); Denis Kleyko (Örebro University & RISE Research Institutes of Sweden, Sweden); Marco Pisani (Sapienza University of Rome, Italy); Amy Loutfi (Örebro University, Sweden); Mauro Olivieri (Sapienza University of Rome, Italy) | Oral |
9:00 AM - 10:30 AM | Session #24: In-Memory Computing II (6 papers) | |||
Time | Paper # | Title | Authors | Type |
9:00 AM | 1571268332 | RINAC: An Energy-Efficient Computing-in-Memory SNN Accelerator Using Redundant Input-Aware Approximate Computing | Wei-Chi Huang and Nathaniel Rensly (National Tsing Hua University, Taiwan); Meng-Fan Chang (National Tsing Hua University (NTHU), Taiwan); Kea Tiong Tang (National Tsinghua University, Taiwan) | Oral |
9:15 AM | 1571268483 | Robust 3-Bit/Cell ReRAM CiM Against Device Non-Idealities by Discrete Shift RTN Suppression and Max Sigma Gaussian Optimized QAVAT | Elliot Yushi Abarra and Naoko Misawa (Tokyo University, Japan); Chihiro Matsui and Ken Takeuchi (The University of Tokyo, Japan) | Oral |
9:30 AM | 1571268509 | MiSS-CiM: Mixed-Signal SRAM Based Compute-in-Memory Systolic Accelerator for Efficient DNN Inference in 2 nm GAA CMOS | Sai Shubham (Indian Institute of Technology, Bombay, India); Harsh Kakadiya, Karan Jaiswal, Megha S and Veeresh Deshpande (Indian Institute of Technology Bombay, India) | Oral |
9:45 AM | 1571273376 | An Energy-Efficient, SCA-Resilient AES Encryption Accelerator Design with Dual-Rail SRAM LIM Macro for Edge Devices | Zhongyu He and Zixuan Shen (Huazhong University of Science and Technology, China); Nay Aung Kyaw (Nanyang Technological University, Singapore); Yuansheng Zhao (Huazhong University of Science and Technology, China); Shirui Sheng (Nanyang Technological University, Singapore); Kwen Siong Chong (Singapore); Bah Hwee Gwee (NTU, Singapore); Chao Wang (Huazhong University of Science and Technology, China) | Oral |
10:00 AM | 1571273548 | Bit Error Tolerance Analysis of High Inference Accuracy AI Using Low-Voltage Floating-Point SRAM-CIM | Tetsuro Mizuno, Seiji Adachi, Yoshinari Wada, Ayane Matsuzaki, Kota Ando, Tetsuya Asai and Takao Marukame (Hokkaido University, Japan) | Oral |
10:15 AM | 1571274583 | A Sparsity-Aware Adaptive Column Replication Framework for Reliable and Energy Efficient RRAM CIM Design | Hyeon-gwon Kim, Seung-Eon Hwang, Dongwoo Lew and Jongsun Park (Korea University, Korea (South)) | Oral |
10:30 AM | Tea Break (15 minutes) | |||
10:45 AM - 11:45 AM | Session #25: Neuromorphic and bioinspired circuits, processors, systems, and their applications II (6 papers) | |||
Time | Paper # | Title | Authors | Type |
10:45 AM | 1571268340 | A Spiking Neuromorphic Architecture for Belief Propagation Decoding of LDPC Codes | Shiang-Yi Chi (National Yang Ming Chiao Tung University, Taiwan); Li-wei Liu and Hsie-Chia Chang (National Chiao Tung University, Taiwan) | Oral |
11:00 AM | 1571268354 | A 1k-Neuron 16bit Edge-FPGA-SoC Reservoir Computing Processor Exceeding 1Msps Input Rate by Zero Skipping for Variable Unstructured Sparsity | Takashi Oshima (Hitachi Ltd., Japan); Yuya Takanuki and Kunihiro Fujisaki (Hitachi Information & Telecommunication Engineering, Ltd., Japan) | Oral |
11:15 AM | 1571268386 | A Firework Inspired Routing Scheme with Single-Packet in-Network Phase Transition for Neuromorphic Systems | Qi Jiang, Fanxi Yang, Lufei Fan, Yi Zheng, Yuhan He, Xun He, Xiaoze Yan and Xirui Peng (Fudan University, China); Hannu Tenhunen (KTH Royal Institute of Technology, Sweden & Ingenium Machina Ltd., Finland); Lirong Zheng (Fudan University, China); Zhuo Zou (Fudan University & KTH Royal Institute of Technology, Sweden) | Oral |
11:30 AM | 1571268466 | PSA: A Pruning-Friendly Sparse-Aware Accelerator for Efficient Sparse Convolution in SNNs | Sanhe Fu and Kainan Wang (Zhejiang University, China); Bo Wang (Singapore University of Technology and Design, Singapore); Aili Wang (Zhejiang University, China) | Oral |
11:45 AM | 1571273646 | RHYME: A Resource-Efficient Hybrid Neuromorphic Hardware System for EMG Gesture Classification | Yuvraj Chowhan (Indian Institute of Technology (IIT) Delhi, India); Chithambara Moorthii J and Anmol Singla (Indian Institute of Technology Delhi, India); Manan Suri (Indian Institute of Technology - Delhi, India) | Oral |
12:00 PM | 1571273836 | L-SPINE: A Low-Precision SIMD Spiking Neural Compute Engine for Resource-Efficient Edge Inference | Sonu Kumar and Mukul Lokhande (Indian Institute of Technology Indore, India); Santosh Kumar Vishvakarma (IIT, Indore, India) | Oral |
10:45 AM - 11:45 AM | Session #26 Special Session: Compute-in-Memory Co-Design: From Emerging Devices to AI Application Optimization II (4 papers) | |||
Time | Paper # | Title | Authors | Type |
10:45 AM | 1571274967 | Explainable and Robust Fake-News Detection System via LLM-Based Credibility and Context Analyses | Tzu-Shuo Peng, Oscal Tzyh-Chiang Chen and Yu Hsiang Hung (National Chung Cheng University, Taiwan); Manh-Hung Ha (Vietnam National University, Vietnam) | Oral |
11:00 AM | 1571275092 | A Hardware-Aware Optimized YOLOv8 for Real-Time Crack Detection at the Edge | Jingjun Sun, Tan Peng, Jiawen Chen, Bashir M. Al-Hashimi and Kai Xu (Kings College London, United Kingdom (Great Britain)) | Oral |
11:15 AM | 1571275722 | Process Variation in Probabilistic Compute-in-Memory: Challenge and Opportunity | Steven Davis, Zephan M. Enciso, Likai Pei and Ningyuan Cao (University of Notre Dame, USA) | Oral |
11:30 AM | 1571275830 | CIMERA: Compute-In-Interconnect and Memory with Reconfigurable Precision for LLM Inference | Yue Jiet Chong, Yimin Wang, Wei Zhang and Xuanyao Fong (National University of Singapore, Singapore) | Oral |
12:15 PM | Lunch | |||
1:30 PM - 5:15 PM | Workshop #1: The 2nd IEEE Chiplet Workshop on Circuits and Systems for Next-Generation AI Computing | |||
Time | Paper # | Title | Authors | Type |
1:30 PM | | Opening Remarks | | |
1:40 PM | CSNAIC-1 | The Carbon cost of Specialization: Rethinking AI Hardware in the Chiplet Era | Tulika Mitra, National University of Singapore | Oral |
2:20 PM | CSNAIC-2 | Scalable 3D Compute-in/near-Memory Systems with Reusable Active Interposer | Chixiao Chen, Fudan University | Oral |
3:00 PM | CSNAIC-3 | Low-Power Wireline/Optical Transceivers for Next-Generation AI Applications | Quan Pan, Southern University of Science and Technology | Oral |
3:40 PM | Tea Break (15 minutes) | |||
3:55 PM | CSNAIC-4 | Progress on Crosstalk Cancellation Techniques for High-Speed Parallel Interconnects | Yuan Du, Nanjing University | Oral |
4:35 PM | CSNAIC-5 | A 16Gb/s/pin 0.51pJ/bit Single-Ended NRZ Transceiver with VDDQ Ripple Compensation for Memory Interfaces | Wenning Jiang, Fudan University | Oral |
| | | | |
1:30 PM - 2:00 PM | Industry talk: HiFloat Data Format Principles, Practices and Standards | |||
2:00 PM - 5:15 PM | Workshop #2: Smart Sensor Circuit and Intelligent Systems for Emerging Applications | |||
Time | Paper # | Title | Authors | Type |
2:00 PM | | Opening Remarks | | |
2:10 PM | S2CAS-1 | Neuromorphic Smart Nose-on-a-Chip with On-Chip Learning for Pulmonary Diagnosis | Hong Chen, Tsinghua University, China | Oral |
2:55 PM | S2CAS-2 | Towards Trustworthy In-Vehicle AI Systems: Dynamic Watermarks for Real-Time Intrusion Detection | Yi (Estelle) Wang, AUMOVIO, Singapore | Oral |
3:40 PM | Tea Break | |||
3:55 PM | S2CAS-3 | Ultra-Low-Power and High Accuracy AI Voice Control Solution: Co-Optimization of Algorithm, Hardware and Toolchain for Edge AI Devices | Jianbiao Xiao, Chip-Artisan Tech. Corp., China | Oral |
4:40 PM | S2CAS-4 | Approximate Computing Circuit Techniques in Efficient Smart Edge Systems - from Biomedical/Vision Recognition to SAR Imaging | Weiwei Shi, Shenzhen University, China | Oral |